CAREER: Nutrition Chatbots: Advancing Nutritional Health and Dietary Counseling Through Human-AI Collaboration

Poor diet is a major contributor to diabetes, heart disease, and other chronic illnesses that affect millions of Americans. Many people struggle to understand nutrition information and health professionals have limited tools for delivering personalized guidance. This project addresses these challenges by developing artificial intelligence systems that work alongside users and decision-makers to deliver clear, personalized, and practical support. The projects uses scenario based interactive learning environments with multimodal data to explore how users understand and act on the offered information. The work advances the national interest by expanding access understandable information for reliable nutrition education, and preparing a skilled workforce at the intersection of computing and health. Educational activities will integrate research into undergraduate and graduate training and engage rural and communities through outreach and open resources. Using chatbots, this project develops and evaluates human-centered artificial intelligence methods for personalized nutrition counseling. First, it investigates scenario based interactive learning environments that help individuals explore tradeoffs and long-term health consequences. Second, it designs a conversational nutrition coaching system that integrates multimodal data including food images, physical activity patterns, and continuous data (i.e., continuous glucose monitoring) to generate adaptive and context aware guidance. Meal photo data will be processed using a large language model to identify personal intake. Data will be collected across three dimensions: activity, sleep, and heart rate. Third, it conducts participatory design studies with decision-makers to ensure that system recommendations align with clinical workflows, ethical standards, and dietary best practice. The research uses mixed-methods including controlled experiments, longitudinal field deployments, and qualitative studies to measure comprehension, trust, behavior change, and clinical usability. Outcomes include new design principles for generative artificial intelligence, open source tools and datasets, and validated evaluation frameworks that support responsible deployment of intelligent systems in preventive health setting. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.